In functional experiments, we have investigated the effect exerted by neurotransmitters released from capsaicin-sensitive primary afferent nerve terminals in the isolated guinea-pig common bile duct. In resting preparations, capsaicin (0.1 microM) produced a quick contraction (45.1+/-4% of KCl 80mM) which was abolished by either atropine (1 microM) or tetrodotoxin (0.5 microM). The tachykinin receptor-selective antagonists GR 82334 (NK1 receptor-selective; 3 microM), MEN 11420 (NK2 receptor-selective; 1 microM) and SR 142801 (NK3 receptor-selective; 0.1 microM) administered separately failed to reduce the capsaicin-evoked contraction, whereas any combination of the three antagonists was effective: GR 82334 plus MEN 11420, 36+/-7% reduction; GR 82334 plus SR 142801, 48+/-4% reduction; MEN 11420 plus SR 142801, 55+/-3% reduction; GR 82334 plus MEN 11420 plus SR 142801, 57+/-5% reduction. Neither the CGRP1 receptor antagonist h-CGRP (8-37) (1.5 microM) nor the P2X purinoceptor antagonist PPADS (50 microM) affected the contractile response to capsaicin. The effect of capsaicin (0.1 microM) was abolished by pretreatment with capsaicin itself (10 microM for 15 min). Human calcitonin gene-related peptide (h-CGRP; 0.1 microM) mimicked the effect of capsaicin on resting preparations (contractile response =28% of KCl 80 mM). In preparations precontracted with a submaximal concentration of KCl (24 mM), and in the presence of atropine (1 microM), GR 82334 (3 microM) and MEN 11420 (3 microM), capsaicin (1 microM) produced a tetrodotoxin-insensitive long-lasting relaxation (45+/-3% reduction of tone, at 4min from administration), which was unaffected by the nitric oxide (NO) synthase inhibitor, L-NOARG (100 microM). h-CGRP (10-50 nM) produced a similar sustained relaxation of precontracted preparations (59+/-4% reduction of tone). h-CGRP (8-37) (1.5 microM) almost completely reversed the relaxations produced by both capsaicin and h-CGRP. Application of electrical field stimulation (EFS: trains of stimuli of 10Hz; 0.25ms pulse width; supramaximal voltage; for 60s) to precontracted preparations produced a sustained, tetrodotoxin (1 microM)-sensitive relaxation (32+/-4% reduction of tone). L-NOARG (100 microM) greatly reduced (69+/-5% inhibition) the EFS-elicited relaxation. A complete reversal of the relaxant response to EFS into a contraction was obtained by administering L-NOARG to preparations in which a functional blockade of capsaicin-sensitive primary afferent neurons had been achieved by incubating the tissue with capsaicin (10 microM) for 15 min. At immunohistochemistry, tachykinin- and CGRP-immunoreactivities (TK-IR/CGRP-IR) were detected in varicose nerve fibers throughout the common bile duct, while TK-IR cell bodies were observed in the terminal portion (ampulla) only. In vivo pretreatment with capsaicin (50 mg/kg; 6-7 days before) decreased the number of CGRP-IR nerves, whereas the TK-IR neural network was apparently unchanged. In conclusion, our data provide functional evidence for the presence of capsaicin-sensitive primary afferent nerve endings in the guinea-pig terminal biliary tract, whose stimulation by capsaicin or EFS produces the release of tachykinins and CGRP. In addition, morphological evidence is provided that the bulk of TK-IR material in the biliary tract is contained in intrinsic neuronal elements, while CGRP in this tissue is of extrinsic origin only. Tachykinins, probably released in small amounts by capsaicin, act by activating receptors of the NK1, NK2 and NK3 type, most probably located on intrinsic cholinergic neurons, which in turn release ACh to produce the final excitatory motor response. The contractile response to capsaicin obtained in the presence of the three tachykinin receptor antagonists could be due to the co-released CGRP and/or to other unknown neurotransmitters. CGRP produces either indirect excitatory or direct inhibitory responses by stimulation of CGRP2 and CGRP1 receptors, respectively.